Expanding Boundaries - Uncertainty and Sensitivity Analysis for the Optimal Design of Distributed Urban Energy Systems – G. Mavromatidis, K. Orehounig, J. Carmeliet [book]

Mavromatidis, G., Orehounig, K., Carmeliet, J.
2016 unpublished
Deterministic model-based design of energy systems assumes perfect knowledge for all the model input parameters. Such a practice entails the risk of sub-optimal designs due to uncertainty that could cause model parameters to deviate from their original values. Thus, the first step towards designing a robust energy system should be understanding uncertainty's effects and its main drivers. Such investigations are facilitated by using uncertainty and sensitivity analysis. In this paper, an
more » ... nty quantification workflow is presented using the energy hub concept as the computational model. Initially, a characterization of uncertain input parameters is outlined pertaining to the buildings' energy demands, and the hub's technical and economic aspects. Subsequently, Uncertainty Analysis (UA) is performed by propagating the inputs' uncertainty through the model in a Monte Carlo fashion to study how the model's output is affected. Finally, Sensitivity Analysis (SA) is performed using the Morris method to screen out unimportant parameters and Variance-based SA to quantify the contribution of input parameters to the output's variance. The framework is illustrated with a case study, a residential urban neighbourhood for which an energy system is designed. The output of this paper allows us to test the model's robustness, understand the extent to which uncertainty actually matters by examining how much the model output varies compared to the deterministic case, and understand the influence and the interactions between the parameters of the model.
doi:10.3218/3774-6_20 fatcat:mi3mt66opbhtpgfm3dcjkams2i